You're right. I guess he does have data out there. It just doesn't seem, from my unscientific eye to mesh with Kai and Woodlands info.
Anyway, to keep the discussion going , here's the quote from the Basic Brewing episode: "It doesn't really generate a lot of new cells, in most people's starters, unless you are doing a fairly big one."
He also kind of scoffs at online yeast calculators, saying that they overestimate.
Here's Neva Parker's slide on yeast growth from her Myth Buster presentation.
View attachment 294487
There's definitely a discrepancy between the pitch rate calculators available to home brewers. Interestingly, I think a lot of the discrepancy comes down to the strains chosen to generate the underlying data. At Omega Yeast, we've undertaken a project to determine the terminal cell density every strain in our collection can achieve under the same conditions in order to provide a strain-specific pitch rate calculator. This might not be immediately apparent to most people, but every yeast strain is capable of growing to a different terminal cell density in the same medium. For example, when started from cells scraped off a plate, the Chico strain (WY1056/WLP001) has achieved cell densities of 148 million cells per mL, which is less than Kai's pitch rate calculator would predict (168 million cells per mL). Some strains, like WY1318, only get to 130 million cells per mL under the same conditions. Some Belgian strains, however, can achieve up to 300 million cells per mL. And Brett upwards of 600-900 million cells per mL. Note that Kai only used WY2042 when generating his data. His numbers for that strain are on par with what we've seen with that strain.
And to complicate matters further, Kai has also shown that stir plate speed impacts cell density achieved by the same strain -- likely because higher stirring speed exposes the growing cells to more oxygen (
http://braukaiser.com/blog/blog/2013/03/25/stir-speed-and-yeast-growth/). We've found his observations on stir plate speed to be accurate. Our numbers line up with his with the "medium" speed results he reports with WY2042.
And yet another factor that plays into the calculations is the amount of cells you start with in your starter. White's observations about less new cell growth in a small starter pitched with a large amount of yeast is accurate. Here is my analogy for that situation. Imagine you need to eat three Twinkies to create a new you. If you're dropped into a room with 12 Twinkies, the prediction is that you can generate 4 new yous for a total of 5. What happens when 12 of you are dropped into a room with 12 Twinkies? If all of you quickly eat a Twinkie, there aren't enough Twinkies to make any new yous, so you remain at 12, despite the consumption of the available nutrients.
So all told, things are a little more complicated than they initially appear given strain to strain variations, stir plate speed, and starting cell density. We've found that for some strains, Kai's calculator does in fact underestimate the cell count and for others, it overestimates.
This goes to show that there's a lot of wiggle room in getting good results when it comes to pitch rate. I generally favor starters for the reasons a lot of people have already stated in this thread -- you know your yeast is viable when you see an active starter and the lag time is shorter, giving any baddies present less chance to cause problems. That said, we're going to make all of our data publicly available when we have it. Our hope is to offer a strain specific pitch rate calculator that incorporates the data we generate for those who wish to get a little more precision in pitch rates, for what it's worth.